In contrast to suggestions that DMS-based benchmarks might not reflect clinical utility, we demonstrate a striking correspondence between VEP performance in functional assays and clinical variant classification.

Explore the full paper for insights into top-performing VEPs.

Traditional benchmarks often face circularity issues, inflating performance estimates. In this study, led by Ben Livesey, we use deep mutational scanning (DMS) datasets from 36 human proteins to benchmark 97 VEPs, introducing a novel pairwise comparison method for fairer rankings.